Flexible production and flexible theory: the case of South Africa

Geuforum, Vol. 26, No. 2, pp. 107-I 19, 1995 Cowright 0 1995 Elsevier Science Ltd

Printedlin &atBritain. All rights reserved 001~718.5/95 $9.50+0.00

0016-7185(95)000224

Flexible Production and Flexible Theory: the Case of South Africa

BEN FINE,* London, U.K.

Abstract: By providing a simple overview of the character of the South African economy as a ‘minerals-energy complex’. the relevance of flexible production to policy making is contested, alongside the presentation of a broader critique of the flee-spec debate. In addition, the intellectual origins of the flexibility approach are identified in terms of dual labour market theory and the simple reversal of the merits of the primary and secondary sectors. This confirms the neglect in policy based on flexibility production both of South African economic realities and the class interests underlying them.

Introduction

There has been an explosion of literature on econ-

omic policy for post-apartheid South Africa. It has emanated from a range of organisations and indi- viduals; it has been constructed from a variety of interests and theoretical and ideological perspectives, with a corresponding mixture of analytical content and quality. From the perspective adopted here, much of this literature has a common deficiency.

Whether concerned with state versus market, export versus inward orientation, high or low levels of wages or state expenditure, getting the prices right or not, there has been a tendency to foist associated analytical predispositions upon the South African economy. In general, these have proved inappropriate because the economy has invariably been inaccurately speci- fied in terms of its characteristics and dynamics.

This is not, however, simply a matter of marrying an improved empirical understanding to an otherwise

*Centre for Economic Policy for Southern Africa (CEPSA), Department of Economics, SOAS, Thornhaugh Street, Russell Square, London WClH OXG, U.K.

unchanging set of policy principles. Any characterisation of the economy needs to go beyond economic structure and dynamics in the narrow sense. For these are dependent upon the class interests to which they are attached. In other words, many policy prescrip- tions are ill-founded, not just because of a skewed empirical understanding of the economy, but also because of a corresponding, even if implicit, misun- derstanding of class relations. This is so even where the attempt is made to promote particular class interests explicitly. The view that such and such a policy will be good for labour or for capital may be limited both by poor empirical foundations and by the way in which labour and its relations to capital are understood.

These general observations are prompted by the wish to provide a more specific response to Rogerson’s (1994) discussion of the prospects of flexible production for the South African economy. This is not simply of academic interest. For the flexibility perspective has inspired the Industrial Strategy Project (ISP) which, closely linked to the Congress of South African Trade Unions, COSATU, has generated detailed empirical and policy research of high quality across a range of industrial sectors as well as for

107

108

strategic issues such as training and research and development.

It is necessary, however, to be acutely aware that, in practice, the individual components of the ISP and the stances of individual researchers have wandered far, and even severed connections from, its initiating analytical framework. To some extent, this has even

been a consequence of ISP’s realisation of the failure of the flexibility paradigm to reflect South Africa’s socioeconomic realities. This judgement is primarily based upon a reading of the draft sectoral reports,’ individual and collective discussion with ISP sector researchers, and the apparent difficulty of the overview documents, ISP (1993), to draw upon the sectoral reports to illustrate its analytical and policy stances. None the less, the focus here is with the theory employed irrespective of whether this has been followed in practice by all or even a majority of those attached to the ISP project.

This runs the danger of constructing a straw figure to knock down. This is even more so, given that much of this commentary will be narrowed down even further to the intellectual origins of the ISP approach. For, although the notion of flexible-specialisation (affec- tionately dubbed flee-spec) on which it is based is little more than a decade old, it has already experi- enced a leap forward in sophistication and scope. Indeed, it has exhibited a degree of analytical flexibility and heterogeneity to match that of the economic developments that are taken as its object. It has encompassed subjects as diverse as changes in technology and the organisation of production, (sub)con- tracting, the size distribution of industry, retailing, distribution, industrial districts, skills and training, retailing, automation, consumer tastes for difference and niche markets, computer-aided management and design, and flexibility within labour markets as well as within production levels, composition and quality, etc. By the same token, the theoretical scope of flexible-specialisation has grown significantly and chaotically. It can be married with notions of post- Fordism and regulation theory to give rise to particular views of the relationship between economics and politics, or it can form a partnership with postmoder- nism to give new meanings to production and, ultimately, consumption. At other times, shifting notions of flexibility and any variety of theoretical fragments can be and have been drawn together

Geoforum/Volume 26 Number 2/1995

under the general umbrella of flexible-specialisation or some other terminology.

In this light, it might be considered a virtue to return to the school’s origins. Indeed, the ISP’s own summary position paper has been through a number of drafts which, at one time, were liable to be amended before they could be distributed and read. This is less

a critical observation than a recognition that the intellectual content of flee-spec has now become open to flexibility in general and especially in response to

empirical heterogeneity and confrontation with South African realities. The purpose here then is to identify two central initiating propositions within the flexibility approach which have persisted even if dis- guised and qualified by subsequent analytical fine- tuning. These revolve around, first, the economic theory employed and, second, the class interests served. But a beginning is made in the next section by addressing some broader issues in the flee-spec debate before revealing the empirical realities of the South African economy for which this debate has tended to induce neglect.

A Few Theoretical Reconsiderations

Despite, as will be seen, presenting a partial view of the economy, Rogerson (1994) has provided a thorough review of the literature on flee-spec, including an account of its deficiencies and its potential implications for South Africa. But there remains a lack of balance in his account. For, whilst noting that “the majority of local manufacturing continues to be rooted upon Fordist principles”, he asserts that flec- spec “represents an important new theoretical concept of significance to the agenda” (p. 12). This is possible only because he essentially treats two arguments as being of equal status irrespective of their

actual or potential empirical weight. As it were, the world could be Fordist or it could be flee-spec (or some combination of the two). Hence, each is potentially of equal importance.

The problem with this is that much of the burden of the literature critical of flee-spec is to point to how much little Fordist production has been replaced, that flexible batch production has become more like mass production rather than displacing it, and that the virtues of flee-spec for commercial success and for the

GeoforumNolume 26 Number 2/1995

benefits of the workforce have been heavily

exaggerated-depending upon a limited number of

success stories, such as the Third Italy regionally and clothing sectorally, which are themselves heavily contested in interpretation.* This is not simply a matter of the empirical validity of appealing to one form of production or another (although the flee-spec school has tended to depend upon isolated case studies and to fail to engage with the presentation of the more extensive contrary evidence on its chosen terrain as well as where it fails to tread). Rather, an idea1 type of

Fordist production is contructed as mass, factory production through deskilled labour of uniform commodities for mass consumption. Partly because of its intellectual flexibility, any variation from this ideal type is construed as conforming to flexible production. In other words, the conceptual balance is tipped in favour of viewing as flee-spec any diver- gence from an ideal type of Fordist production. Con- sequently, to defend analytically the importance of mass production, factory techniques, tendencies to deskilling, etc. is not to be blind to the pervasive significance of new technology and organisational change. Rather, it is to reject the idea that there is only one immutable form of Fordist production (and that all production must be Fordist for mass pro-

duction to be in the ascendancy).3

Flexible specialisation, then, holds to a view of the world in which the economic and social conditions associated with Fordism or mass production are in crisis and are possibly drawing to an ultimate close. In the future. building on existing trends, there is an alternative path of development associated with smaller scale batch production, satisfying more specialised demand and offering more skilled and varied

work.

But, as has now been recognised in the prolific debate over flee-spec, even amongst many flee-spec sym- pathisers, a simple dichotomy between Fordist and post-Fordist forms of production is unacceptable. As

Gertler (1988, 1989, 1992) argues from a relatively critical stance to flee-spec, ‘flexibility’ itself needs to be unbundled into a range of separate components- not least those mentioned above in terms of the increasing sophistication, scope and flexibility and heterogeneity of flee-spec itself as a concept. It is not necessarily the case that the separate components will always have shifted in the same direction, towards or

109

away from the new forms of production. Indeed, it is

apparent that flexibility in one component of an economic system may require rigidity in another component-a large, inflexible dedicated machine, for example, can embody capacity for flexibility in product mix. This is obviously the case for the car industry, for example, which is why it can be claimed by both sides in the debate over flee-spec.

For this reason, as mentioned above, those opposed to the flee-spec stance are reasonably drawn to pointing out the limited extent to which all of its features are empirically realised in all but a few select industries and localities. By the same token, especially as illustrated by Schoenberger (1989) in debate with Gertler. those favouring flee-spec argue that not all of its elements need to be in place for its presence to be empirically verified. There only has to be a movement towards them, Schoenberger adding a further appeal from regulation theory to systemic shifts in the regime

of accumulation.

Gertler reasonably concludes, then, that the theoretical debate has only rendered the empirical questions more impenetrable. Given there is no rigid distinction between Fordism and post-Fordism, how is it possible to identify a shift from one to the other, whether tied to a regime shift or not, especially if acknowledging that flexibility is intrinsically subject to contradictory and opposing developments across the separate components that mark the Fordist/post- Fordist divide? When is an industrial district to be distinguished from a spatial conglomeration of producers; what is a small as opposed to a large firm; when is flexible technology at the leading or lagging edge, signifying a continuity or rupture with the past, overhauling or overthrowing what has gone before?

Apart from these theoretical/empirical conundrums, Gertler points to a number of factors that have been insufficiently incorporated into the debate; how are inter-sectoral boundaries established and restruc- tured. what is the significance of gender and class relations, national systems of innovation, and the reconstruction of space in the light of the pace and nature of technological change? No doubt these could be added to, or developed within, a continuing debate even if at the expense of continuing erosion of the Fordist/post-Fordist divide. There is, however, an alternative option, one that is both readily sugges-

110

ted by the increasing eclecticism of the flee-spec debate but which breaks with it entirely. Rather than basing theoretical and empirical analysis upon the putative dichotomy which can only be sustained by increasingly complex analytical acrobatics, why not abandon this as a starting point?

Of course, this raises the problem of what to put in its place. But this is a false problem since the flee-spec

claim of discovering previously unanalysed, and now leading edge, economic developments is itself totally fallacious once the simple dichotomy between For- dism and post-Fordism is rejected. All the components that make up the diode have long been the subject of theoretical and empirical investigation. The debate may, rightly or wrongly, have brought some of these more to our attention than previously and have prompted theoretical innovation and flexibility, whether in support of the flee-spec stance or not. It is arguable, however, that whatever the role of the debate previously in stimulating intellectual progress, its ‘market’ niche is now exhausted and it has

passed its sell-by date.

Quite apart from the conundrums identified by Gertler, this conclusion is confirmed by a particular way in which empirical evidence is now being handled theoretically. Corresponding to the different, empirically identified, aspects of (non-)flexibility, it is possible to argue that they constitute contradictory tendencies. These theoretical constructs, in turn, can be used to ‘explain’ the empirical trends from which they have been induced. Thus, where flexibility is

contradicted empirically, this can be interpreted as an impediment to, or even reaction against, leading edge development.4 In other words, evidence either in support of or against either stance can be interpreted favourably in view of contradictory tendencies. Gertler is, himself, notably guilty in this respect, suggesting the concept of ‘glocalization’ to signify the contradictory interaction of globalization and localiz- ation.

The purpose here is not to point the accusing finger of non-verifiability. Rather, it is to question for the last example as well as more generally for the flee-spec debate, whether the dichotomies being drawn are analytically appropriate. Perhaps the proof of the pudding is in the eating. What follows in the next section attempts to show how the imposition of the


flee-spec analytical structure upon South African is to eat the wrong pudding, thereby serving as an obstacle to understanding.

The South African Economy and the Minerals-

Energy Complex

In the event, Rogerson (1994) provides only an ex-

tremely limited and general empirical specification of the South African economy in assessing the scope for

flexibility. It is also one that is totally misleading where it does have some purchase. He refers explicitly to Gelb’s (1991) notion of racial Fordism and to Black’s (1991) notion that there has, over the past two decades, been a reversion to primary production. The attraction of racial Fordism as an analytical category is that it explicitly addresses a unique aspect of the economy, its peculiarly longstanding and oppressive racial inequalities. But at a more general level, it suffers from characterising the South African economy by what it is not (i.e. a non-racial Fordism). More particularly, it is unable to explain the particu-

lar trajectory of the South African economy-why strong in primary commodities but not manufacturing-and ultimately rests upon undercon- sumptionist propositions (black wages are too low to sustain growth) which, whatever their merits, are theoretically independent of Fordism and racism.

In short, racial Fordism represents a simple wedding of a regime of accumulation called up from regulation theory, ‘Fordism’, to an equally simple empirical

understanding of apartheid as ‘racial’. There can be no presumption of the prospect of a non-racial post- Fordism by analytical extrapolation from the crises of both apartheid and Fordism. But the notion that flexibility is an alternative to reversion to primary production is equally flawed since it presumes that the significance of primary production for the South African economy had previously been discarded along the path of industrialisation. This is disputed here through a fuller and competing characterisation of the economy.

Indeed, in previous work,5 it has been argued that at the core of the South African economy exists what has been termed the minerals-energy complex (MEC).6 Only by unravelling its mode of operation and its intimate relations to the state and finance is it


possible to address the issue of the weakness of manufacturing, which other analyses take as their

starting point, and the prospects for different policy strategies.

But what is meant by the MEC? First, in identifying the MEC as lying at the core of the economy, there is at its own core those sectors associated with minerals and energy in the narrow technical sense. If we begin with coal,’ it is responsible for 80% of the country’s primary energy needs. It also provides for 20% of the country’s exports (one third of non-gold). But very little coal is used for direct energy needs. The vast majority is either converted into electricity (over 50%) or into oil (over 30%). The South African economy is uniquely dependent on electricity and is uniquely electricity-intensive with levels of consump-

tion per capita comparable to those of the U.K, despite limited domestic consumption by the majority of the population. This is primarily due to its use in mining and mineral processing, this accounting for 40% of consumption. Even though manufacturing accounts for a slightly greater proportion, the industries concerned are closely related to, indeed form part of, the minerals-energy complex with heavy use in a small number of plants in engineering, iron and steel and base metals and chemicals. Thus, coal is produced to generate electricity which acts as a major

direct and indirect input to the production of gold. In this light, it is not surprising that some have estimated gold as generating 40% of GDP through its direct and indirect effects, Lombard and Stadler (1980). Although gold’s importance has declined over the

past decade, other minerals have been of increasing importance (which is what prompts the reversion thesis).’

The relevance of flee-spec to these core primary sectors has to be seriously questioned. Even within a framework imposed along the lines of a Fordist/post- Fordist distinction, the vast majority of this core of the MEC is unambiguously dependent upon Fordist- inclined production. All of mining is hugely capital- intensive and, despite heterogeneous mining conditions, is concerned with digging up and processing massive quantities of ore. For gold, it is a matter on average of extracting as little as five grammes of the precious metal from each tonne of crushed and pro- cessed ore, itself removed from as deep as five kilo- metres below the earth’s surface.

0 Non-MFC manufact.

:924 1930 1936 1942 1948 1954 1960 1966 1972 1978 1984 1990

Figure 1. GDP contribution minerals-energy complex, 1924-1990.

Source: Union Statistics for Fifty Years (1960), South African Statistics (1990), IDC (1992).

Coal is mined using state-of-the-art technology in both underground and surface mines on an equally

massive scale. As a product, coal comes in many qualities and is not a homogeneous output. This leads it to be washed, prepared and transported for use through economic activities that might well be de- scribed as archetypal Fordist. Similar considerations apply to other mining sectors, concerned with extracting and processing huge quantities of the earth’s mass.

The main use of coal is in power stations, once again, employing the most advanced large-scale technolo- gies, with the most recent of stations combining six

generating ‘sets’ and, hence, six times the size of that typical of the U.K. South Africa produces more than half of the whole continent’s electricity supply.’ Otherwise, export of coal is dependent upon ‘Fordist’ transport and, unique to South Africa, coal-to-oil conversion is on a massive scale, providing for 50% of domestic needs.

Second in specifying the nature of the MEC, whilst the relative importance of its core sectors has declined, this iS only because of the more than propor- tionate increase in large-scale manufacturing

immediately downstream from the core-as in heavy chemicals, specifically integrated with the coal-to-oil

conversion facilities, and other mineral and metal processing, for which the previously state owned mass steel producer, ISCOR, has been central. Far from ever having declined, the weight of activities in and around the core mineral-energy sectors has even increased, as shown in the Figure 1.” This demon- strates how important the MEC has been, and continues to be. once it is recognised to incorporate core

112

sectors and those immediately downstream from them. The central and growing role of these sectors

taken together has been overlooked in the literature because of an uncritical acceptance of an inappropriate division for statistical purposes between primary sectors and manufacturing. Much of the MEC activity has been classified as belonging to a general category of industry, even though it is much more closely aligned, as in heavy chemicals, metal processing and much heavy engineering, to the core MEC sectors.

If, in addition, industry is not itself carefully dis-

aggregated, a second fallacy arises-the view that South Africa has predominantly industrialised, however successfully, around import-substitution for consumption goods which have, indeed, enjoyed considerable protection. But the growth of manufacturing has been too readily, and incorrectly, associated with the heavily protected consumer goods industries. To the extent, then, that flee-spec strategies depend upon departing from an ISI-led industrialisation around consumption goods, South Africa has yet to arrive satisfactorily at the first staging post from which flee-spec takes its point of departure. Indeed, the process of industrialisation has moved more around primary production than consumption goods.

Consequently, whatever the merits in general of an industrial strategy based upon flee-spec, the notions that South Africa has industrialised and that its industrialisation is consumption-sector based (as in notions of racist-Fordism where only the white population’s needs for mass consumerism are met) represent false starting points. A recognition of the continuing and growing significance of the MEC is a dampener on flee-spec strategies with their inappropriate emphasis upon niche markets and flexible production.

A third feature of the MEC is that it is bound to a particularly concentrated corporate ownership structure. Effectively, five large-scale corporations run mining and, where they are not dominant across the MEC, huge public corporations have been central, as in electricity, steel, transport, etc. In addition, the mining houses are at the centre of a much wider set of activities than of mineral production alone. There is an extensive pattern of interpenetrating directorships and ownership of shares and a highly concentrated ownership of other companies within the economy as


a whole and abroad. Anglo-American, for example, is well-known to own over 70% of the shares on the Johannesburg Stock Exchange and to be the single largest foreign investor in the United States. Further, the conglomerate structure of the MEC depends upon, and is reflected in, a close integration with South Africa’s highly developed financial insti-

tutions. For each mining house, there is not only a corresponding holding company for its industrial activities but also a matching set of financial institutions. Thus, the corporations effectively hold economic power, and have demonstrated little or no change in strategic policy making, preferring to seek

continued subsidies for large-scale MEC type mega- projects and to engage in (illegal) capital flight. ANC policy has shifted considerably from challenging this economic power, even through the leverage that it has in the control of State corporations. Instead, State vice-president Thabo Mbeki has proposed privatisation as a means of financing the Reconstruction and Development Programme (RDP, 1994).”

It is against this broadly brushed economic landscape that prospects for South African flee-spec have to be

judged. Within the horizon must, however, also be included levels of access to formal, waged employment not far above 50%, and the lack of means of satisfying basic needs for the majority of the population. Significantly, Rogerson looks to the potential for flee-spec arising out of diversification from the sanctions inspired armaments industries, and from sectors such as electronics, computers, biomedical equipment and security systems (and the ubiquitous clothing industry which is of considerable importance to the South African economy). Geographically, the Midrand and Cape Town region are noted as potential growth points for flee-spec; the effect would be to consolidate what are already heavily skewed patterns of industrial and economic activity.” In short, a flec- spec strategy would fail to address unemployment on a significant scale, the current structure and composition of economic activity, corporate power, and the means of providing basic needs for the majority of the population.

Flee-spec’s Intellectual Origins

How is it that flee-spec can be so fervently pursued even where it confronts such inappropriate con-


ditions as those to be found in South Africa? Perhaps

the answer is to be found in the intellectual appeal of

new ideas despite the weight of contrary evidence. For this reason, it is worth uncovering what has been an initiating and persistent theme within the favourable literature. In this respect, the classic statement is to be found in Piore and Sabel (1984).

For the fundamental position adopted here to that

work is that the flee-spec school, even if diluted by sophistication and a range of other considerations, has retained at its core a faithful reproduction of an earlier theory, that attached to dual labour markets. This theory was advanced in the early 1970s and argued that the labour force is divided into two strata, the primary and the secondary, in correspondence to an industrial structure based on mass production and casualised production, respectively. Those in the core or primary sectors enjoyed well-paid, secure jobs with career ladders. Those in secondary or peripheral

labour markets suffered poor wages and working conditions and the immanent threat of unemployment. This theory was developed to explain black riots in the U.S. city ghettos-as a consequence of employment deprivation. It has since blossomed into a much more sophisticated understanding of the structure and causes of labour market segmentation, just as the flee-spec school has incorporated more considerations. Dual has given way to segmented labour market theory, with many more segments and explanatory factor.13 What the flee-spec school has done is simply to reverse the ranking of the primary

and secondary sectors, with the latter potentially providing the key to economic progress and favourable working conditions. Indeed, this strong link between dual labour market theory and flee-spec is not difficult to discover, for it stands out in the person of Piore who co-authored the classic work both on dual labour markets with Doeringer (Doeringer and Piore, 1971) at its (modern) outset and on flee-spec with Sabel subsequently.‘4

But the parallels between the two positions still have to be identified. Begin with the economic theory involved. First, there is the question of industrial structure. Of necessity, flee-spec offers a duality between mass production and flexible production, and these are core concepts. It is argued that mass production won out in the struggle for ascendancy at

113

the end of the nineteenth century, although this was not an inevitable outcome if economic and political conflicts had been resolved differently. But once halfway established, whether as a paradigm to be copied or by the inertia embodied in fixed capital and institutions, the mass production model’s dominance

was guaranteed-at least for the better part of a century.15

Even so, flee-specs fundamental viability is demonstrated by the otherwise presumed anomalous sur- vival of small-scale and flexible methods of production. These have themselves been given a new

lease of life by the crisis of Fordism and by the potential opened up by new technology. The historic choice between mass and flexible production is to be offered once more. Hence, the Second Zndustrial Divide is the title take by Piore and Sabel (1984) and the divide re-opens before us.

Significantly, the division between mass and flexible sectors is not determined by economics and certainly not by technology alone. Here the parallel is with segmented labour market theory in so far as the latter depends on contingent supply and demand factors, in which sexism and racism, for example, assume a prominent role. l6 For flee-spec, more emphasis is given to the more orthodox factors of social theory, such as trade union organisation, the community and government support (predominantly at the excep- tional local level for flee-spec to prosper, since central and normal local governments serve the needs of mass production). l7

None the less, there is an economic theory at the core of flee-spec and it is identical to much that is to be found in segmented labour market theory. This is explicitly recognised by Sabel(l982: 34) who refers to Piore’s earlier work. Central is the division of demand into a stable and a variable component with the allocation of each, respectively, to the mass and flee-spec sectors. All that is now added is that the former of these is restricted by the extent of the market, Piore and Sabel (1984: 55), whilst the latter can become more secure by frequent innovation to satisfy specialised market niches, Piore and Sabel (1984: 27). Stable demand around mass production/ consumption is now insecure; unstable demand is more secure through innovation, etc. to serve market niches.

114

But what is it that has opened up the second industrial divide? At a theoretical level, as we have seen, in principle, the relative merits of mass production and flee-spec are presumed to be ever present. Mass production’s historically contingent success lies purely in its initially gained lead which ultimately moulded economic and social institutions to its advantage. Otherwise the choice is between dedicated low unit cost, high fixed cost unskilled mass production and high unit cost, low fixed cost, skilled,

specialised production. Much less is made of the impact of new technology in swinging the balance in favour of flee-spec than might be expected in view of later literature.

Accordingly, a search is made for the economic and social causes of the decline of Fordism and for the

social factors promoting flee-spec. The economic crisis and slow down of the seventies certainly signals for the flee-spec school that mass production is down if not necessarily out. Explanations in terms of ‘shock factors’, such as oil crises, etc., are viewed with suspicion and more linked to the timing of the crisis, Piore and Sabel (1984: 168). More systematic and secular within the economic plan is the exhaustion of demand, for “the crisis of the last decade is a crisis of

underconsumption, rooted in the saturation of core markets for consumer durables”, Piore and Sabel (1984: 252).

Of course, within political economy, an unresolved issue has always been whether underconsumption is a cause rather than a symptom of crisis. Certainly the one must be accompanied by the other and, similarly, mass producers are expected to form cartels, seek out guaranteed markets and to suffer excess capacity from time to time. Such symptoms cannot be taken to

be an adequate explanation for, and characterisation of, the crisis of Fordism. But flee-spec theory also examines the cost side of mass production. Here, it is argued that ‘hidden costs’ have imperceptibly crept up on transnational and multiplant producers. These are predominantly concerned with management response to the underlying and determining conditions of demand. Thus, for Piore and Sabel (1984: 200), there are problems of managing dispersed sites of production, maintaining inventory and quality control, bending standardised products towards frag- menting tastes and in dealing, more generally, with

Geoforum/Volume 26 Number 2/1995

instability when constrained by large-scale, dedicated, production methods.

In the case of labour, mass production yields a systemic compromise in which workplace control and wage costs are stabilised in return for security of employment, Piore and Sabel (1984: 64). Both the

organisation of work and the levels of wage costs become increasingly inflexible relative to the market

demands that are placed upon them, although the whole of this argument depends upon setting aside the superior productivity (gains) associated with mass production (for which the notion of stable wage costs becomes problematic). Again, it is questionable whether pointing to the notion of inflexible labour in wages and conditions is a recognition of a (management perspective) symptom of the crisis rather than the identification of a fundamental cause.

But whether in looking to the labour or the product markets, the transnational corporation is perceived to be the, increasingly pressurised, microeconomic regulator of the economic system. This represents a welcome break with orthodox neoclassical economics, for which transnationals might just as well not exist. At the macroeconomic level, however, the orthodox notion of Keynesian demand management,

coupled with welfarism, is adopted as the regulator, increasingly powerless to satisfy the economic and

social demands placed upon it. At the risk of being tediously repetitive, it must again be observed that the crisis of Keynesianism is more a symptom than a cause of crisis and, for that reason, a common ingredi- ent to every analysis of the last decade or more.

As such, this not only reflects limited analytical per- ception, it also seriously understates and distorts the economic role of the state. For its economic policies have been equally concerned with microeconomic regulation, in supporting and promoting the model of

transnational and large-scale corporate growth- through industrial and other policy measures (not least public ownership)-and these are arguably as significant to mass production as manipulation of aggregate demand. No doubt, such points are to be warmly embraced by the flee-spec school as supporting their view that mass production has an institutio- nalised advantage. But it leaves open the unanswered question of why such advantages should have proved insufficient to sustain the Fordist model. It also serves

GeoforumNolume 26 Number 2099.5

to accentuate the (false) idea that Fordism involves more of a separation between economy and society than flee-spec so that the latter becomes a novel model with potential derived from the evaporating boundaries between economic and social organisation.

In short, the crisis of Fordism is perceived as a structural crisis of labour, corporate and social organisation which has been brought about by the inflexi-

bility of the system to respond to volatile, faltering and increasingly sophisticated demand. The mirror image of this is to be found in conditions favourable to

flee-spec. Apart from the demand conditions and new technology, the most important factor is the con-

tingent political conditions that have been created, predominantly at the regional or local level, to which progressives have turned, following curtailed or disil- lusioned activity at the national level. The stereotype is provided by Third Italy, Mondragon and, as some would have it, London, with particular emphasis on fashionable, flexible demand goods such as clothes and furniture. ”

The ideology of the second divide depends then on a ready identification with each other of different in-

stances of the paradigm or principle of the craft worker, forged not only across space and sectors, but also historically to create a connection with the first divide. In addition, this homogenising of flee-spec production into a single principle is heavily romanti- cised. Thus, for Sabel (1982: 107/g), we have the “fusion of conception and execution”, “with a bright idea”, “he begins to attend night-school clases”, “the ability . . to make virtually any tool in his shop do

new tricks”, and each firm is “jealous of its auton-

omy”, “overproud of its capacity”, “linked to the

collective efforts of the community”, whilst the “relations between firms resembles the collegial relation between good doctors, good lawyers, or good university teachers.”

Such paradises are few and far between, certainly as rare as the good professionals last named. Flee-spec is also well aware that the secondary sector and/or labour market can become the site of subcontracted, sweated work as the source of flexibility for the corporations reeling under the crisis of Fordism. Indeed this trend is a symptom of the crisis. Only progressive forms of local economic and political

115

organisation to create industrial districts/

communities can offer any guarantee against this. On the other hand, flee-spec can develop within the corporations themselves, by reorganisation of work and marketing. The idea is that production is geared towards small batch production, with more flexible fixed capital and a greater diversity of skills and job content for teams of workers enjoying greater job autonomy and control. Examples offered include

mini steel mills, chemicals and machine tools, Piore and Sabel(l984: 205), but the most stunning convert is to be found within the car industry, with the decline in the strategy for a world car and greater emphasis on offering model selection, Katz and Sabel (1985).

This hardly conforms to craft production (and con- centration within the car industry has increased since the seventies as transnationals such as Chrysler and British Leyland have fallen behind into a lower lea- gue). As Katz and Sabel (1985: 298) fully recognise:

The emerging strategy is to group automobile components into modules, each produced in high volume, and then combine the modules into a wide variety of finished products the entire manufacturing set-up becomes strikingly flexible by the industry’s previous standards.

This conclusion is not analytically obvious. The dis- tance between the car industry and the flee-spec industry of the first divide and even of small-scale production of the second divide constitutes a gaping chasm which is hardly bridged by some abstract

notion of their sharing a common response of flexibility to market uncertainty.

Thus, as has been so often emphasised in the subsequent literature, flee-spec in the modern world

seems way-off the mark both analytically and empirically. But it should not be dismissed too easily and purely on academic grounds. For it does. despite its Utopian vision for the future,‘” support more im- mediate policies, certainly as far as industrial relations and economic intervention are concerned. Underlying flee-spec proposals for each of these is the idea that compromise can be reached between capital and labour so that each can benefit from the high quality, variable production and the enhanced and varied work. Conflict between the two sides of industry is not absented, but it is required that labour make concessions, particularly at the level of national

116

agreements so that the plant level workforce can bargain flexibly with management even if within a broader framework of higher level bargaining, Katz and Sabel (1985: 314):

Unions will have to exchange the rights to impose uniform conditions for rights in decision making . the unions will have to find ways of tying the interests 01 particular companies to the interests of their industry and even of the economy as a whole. . Under extra- ordinarily favourable circumstances, national unions might even help structure the strategy of whole industries . . . and have a say in macroeconomic decisions.

Although derived from different arguments and considerations, the flee-spec approach reaches similar conclusions to those put forward by cautious social democrats who seek limited state intervention and wage increases.‘” Capital and labour need to reach an

accord in order to benefit mutually from the attached productivity or commercial gains. ISP has adopted this stance with a particular emphasis, in acknowl- edgement of South African conditions, of the importance of education and training.

However, exactly the same reservations apply to ISP perspectives for South Africa as have been identified throughout the course of flee-specs evolution. How far does it apply to the South African economy; even to the extent that it does, what guarantees are there of progressive forms of flexibility; and, in essentially leaving aside the issue of large-scale capital, will it not merely support, however successfully and however much in conformity with its ideal types, the emerg-

ence of a ‘rent-seeking’ strata of small-scale capita- lists?

The purpose of this extended discussion of flee-spec is less to assess its salience as to illustrate how the class content of policy perspectives often occupy a secondary position even where it is explicitly addressed. The ISP does seek to address and promote the interests of labour even if in compromise with capital. But its particular preoccupations, with their origins in flec- spec analysis rather than South African realities, imply an undue neglect of the nature and power of large-scale capital. Specifically, there is the danger of overlooking the continuing pressure for the promo- tion of the MEC through mega-projects (the main agenda) even as these are veiled by simultaneously

GeoforumNolume 26 Number 20995

publicising the advance of small-scale enterprises? By the same token, strategies based upon the capa- bilities of the ‘Fordist’ MEC tend to be precluded. It is hardly surprising that Rogerson is able to observe that, “policy interest in flexible production extends across the political spectrum in South Africa” (1994:

8). For it is a strategy which is increasingly malleable, supporting small business without threatening large- scale capital, and which has increasingly been pre- sented in neutral terms as good for the economy or whatever rather than, as initially, a strategy for labour however well-conceived.** This increasing

‘neutrality’ of flee-spec proposals deepens the closer they approach policy formulation and implemen- tation, especially when business interests have to be won over.

Concluding Remarks

At an analytical level, the intention here has been to move beyond the flee-spec debate. Further, even so, it has been to indicate how inappropiate the flee-spec approach is to the empirical and policy context of the South African economy’s dominance by its MEC. Last by uncovering the intellectual origins of increasingly flexible flee-spec theory in the inversion of dual labour market theory, the potential dangers are re- vealed of flee-spec policy making both in overlooking the power of large-scale capital and in promoting the interests of small business, whatever their merits, at the potential expense of working class organisation, militancy and wages and working conditions.

What is the alternative? It is not the intention here to outline alternative policy proposals, but they would be based on the follqwing, far from exhaustive, considerations drawn from South Africa’s political and economic realities.*”

1.

2.

South Africa has considerable infrastructural capacity but a highly racially skewed provision. Priority must be given to public sector pro- grammes for such basic needs as health, education, housing and electrification.

Industrial policy must be addressed in terms of redressing selectively the bi-modal structure of


3. The economic power and effects of conglomerate

capital must be confronted directly rather than avoided through what will at best prove to be mildly supportive policies for small business. The latter’s potential is not to be denied altogether but

it must be placed in perspective, including the

brighter prospects the more the core of the economy prospers.

In short, the constitutional settlement has already

manufacturing, around the two extremes of MEC core and consumer goods, by developing integrated capability in intermediate and capital goods.

considerably weakened the powers available to the Government of National Unity. This is being reflected in the declining ambition and increasing mar- ginalisation of the RDP.24 Blowing the trumpet of flexible production can only promote and conceal

these unfortunate developments in attempting to come to terms with the heritage of apartheid.

Acknowledgements-The support of the Economic and Social Research Council (ESRC) of the UK is gratefully acknowledged. The work was funded by ESRC award number RO00232411 for a grant to study the South African minerals-energy complex. Thanks to anonymous referees for comments on an earlier version.

Most of these are to be revised and published by the University of Cape Town Press. The recent special issue of World Development, edited by Humphrey (1995) and devoted to flexibility in developing countries, is remarkable both for failing to recognise how limited is its scope of application and, on which see later, for neglecting the role of underlying economic interests and their political expression. The latter is primarily seen as affecting the form in which flexibility is adopted, conveniently allowing for further analytical flexibility in empirical interpretation. For an outstanding critique of the binary opposition between Fordism and flee-spec, and its analytical imposition upon the development of food systems, see Goodman and Watts (1994). Fine (1994) criticises a similar method in food systems theory, where empirical trends are treated as tendencies whose contradictions are ultimately perceived to require structural change. See the debate in the same place that this inspired as well as Fine et al. (forthcoming).

5.

6.

This account represents a drastic summary of Fine and Rustomjee (1995), where sources and references to the literature are fully cited. See also Fine and Rustomjee (1992) and Rustomjee (1992). Interestingly, Kaplinksy (1995), a director of the ISP, seems to accept the relevance of the MEC as a basis on which to analyse the South African economy. His own explanation for its influence, adjudged to lead among other things to neglect of investment in labour- intensive sectors, ultimately rests on ‘political’ factors. But, as has been made clear in the work to which he himself refers, these political factors involved the crea- tion over the post-war period of Afrikaner large-scale capital and its integration with large-scale ‘English’ capital. Thus, the MEC and the patterns and levels of investment are a consequence of economic factors even if they have been supported politically. Significantly, supporting our introductory comments on the increasingly acknowledged mismatch between the flee-spec stance and South African realities, Kaplinsky’s analysis also only makes token reference to the flee-spec approach and, then, only as one possible industrial policy response to the changing political circumstances in the future. The approach seems to have nothing to offer in explaining the past and the continuing economic interests that it has fostered.

7.

8.

9.

10.

11. 12.

13.

14.

15.

16.

For a critical review of the South African coal industry, see Fine (1992). For the most recent information on mining and energy, see the reports of the Department of Mineral and Energy Affairs. See Nattrass (1995) for a recent study of the South African gold industry. See the annual statistical reports of Escom, the state owned electricity utility, for a comprehensive account of the empirical significance of electricity within the South African economy. For details of the division between MEC and non-MEC manufacturing, see Fine and Rustomjee (1995). For a critique, see Fine (1995b). The point here is less to refute the implicit flee-spec analysis of these sectors, which can only be done in detail. but to signal how limited is the weight even of those sectors that are targeted, especially for the trans- formed needs of the economy. Rogerson also raises the car industry which is already universally recognised to be far too fragmented. What follows is drawn from an appendix to Fine (1987) which offers a thorough critical review of segmented labour market theory. It is significant that the intellectual origins of flee-spec could have been identified so early and as a corollary of an earlier tradition in labour market theory (rather than as a pure response to the demise of Fordism. however well understood in theory and realised in practice). As far as the reversal of market structure is concerned. this is also observed by Gertler (1992: 265). For the flee-spec history reported in this paragraph, see Sahel and Zeitlin (1985). In Sahel (1982). however, workers are invariably ‘he’. etc. and this may in part reflect the central role of craftsnzen in flee-spec analysis, on which see later. The role of ethnic groups is more prominent since they are more liable to form a community for economic as well as for social purposes.

17. Thus. Piore and Sabel (1984: 275) (incorrectly) assert:

117

118 Geoforum/Volume 26 Number 2/1995

18. . . . ends, and where economic organisation begins.

These areas and sectors are cited so often as to have become cliches of the flee-spec school. For a critique of the notion as applied to the Greater London Council’s industrial strategy, see Nolan and O’Donnell (1987), for example. The story of the GLC’s ‘interventions based on flee-spec has yet to be fully and properly told. The present author had some involvement and would emphasise how difficult it was to find appropriate investments and how disastrous those were to prove that were made.

19. Thus, in Piore and Sabel (1984: 276/79):

A shift away from mass production would restore the neoclassical equilibrating mechanisms . . the risks of investment will be generally reduced . . it would be possible to maintain full employment largely through monetary policy . . there would be much less danger of shocks to the economy touching off inflationary surges . . Similarly, flexible specialisation is better able to accommodate fluctuations in exchange rates and commodity prices , . . flexible specialisation and mass production could be com- bined in a unified international economy. In this system, the old mass-production industries might migrate to the underdeveloped world, leaving behind in the industrialized world the high-tech industries . To the underdeveloped world, this hybrid system would provide industrialisation.

20.

21.

22.

23. 24.

Otherwise, the alternative is a resumption of Fordism on the basis of a multinational Keynesianism to provide adequate demand. For a critique in the South African context, see Fine (1995a) for example. This currently continues to be the line adopted by the South African quango, the Industrial Development Corporation (IDC, 1992), in a remarkable reprise of its role under the apartheid system. Rogerson’s only mention of the class basis for tlec-spec is a brief warning against the dangers of sweating. For further discussion, see MERG (1993). See Adelzadeh and Padayachee (1995) for a critique of the shifting content of the ANC’s RDP.

flexible specialisation works by violating one of the assumptions of classical political economy: that the economy is separate from society . . . By contrast, in flexible specialisation it is hard to tell where society

References

ISP (1993) Meeting the global challenge: a framework for industrial revival in South Africa, In: South Africa and the World Economy, P. Baker, A. Boraine and W. Krafchik (Eds). David Philip, Cape Town.

Kaplinsky, R. (1995) Capital intensity in South African manufacturing and unemployment, 1972-90, World Development, 23, 179-192.

Adelzadeh, A. and Padayachee, V. (1995) The RDP White Katz, H. and Sahel, C. (1985) Industrial relations and Paper: reconstruction of a development vision, Truns- industrial adjustment in the car industry, Industrial Re- formation, 25, 1-18. lations, 21.295-315.

Baker, P., Boraine, A. and Krafchik, W. (Eds) (1993) South Africa and the World Economy in the 1990s. David Philip, Cape Town.

Black, A. (1991) Manufacturing Development and the Economic Crisis: a Reversion to Primary Production, In: South Africa’s Economic Crisis, S. Gelb (Ed.). Zed Press, London.

Doeringer, P. and Piore, M. (1971) Internal Labor Markets

Lombard, J. and Stadler, J. (1980) The Role of Mining in the South African Economy. Bureau for Economic Policy and Analysis, University of Pretoria.

MERG (Macroeconomic Research Group) (1993) Making Democracy Work: a Framework for Macroeconomic Policy in South Africa. CDS, Cape Town.

Nattrass, N. (1995) The crisis in South African gold mining, World Development, 23. 857-868.

and Manpower Analysis. Heath Lexington Books, Lex- ington.

Fine, B. (1987) Segmented Labour Market Theory: a Critical Assessment, Birkbeck Discussion Paper in Economics, no 87/12, reproduced in shortened forms as Thames Papers in Political Economy, Spring, 1990.

Fine, B. (1992) Total factor productivity versus realism: the South African coal mining industry, South African Jour- nal of Economics, 60,277-292.

Fine, B. (1994) Towards a political economy of food, Review of International Political Economy, 1,519-545.

Fine, B. (1995a) Politics and economics in ANC economic Policy: an alternative assessment, Transformation, 25, 19-23.

Fine, B. (1995b) Privatisation and the RDP: a critical assessment, Transformation, forthcoming.

Fine, B. and Rustomjee, Z. (1992) The political economy of South Africa in the interwar period, Social Dynamics, l&26-54.

Fine, B. and Rustomjee, Z. (1995) South Africa’s Political Economy: From Minerals-Energy Complex to lndustria- lisation? Hurst, London.

Fine, B., Heasman, M. and Wright, J. (forthcoming) Con- sumption in the Age of Affluence: the World of Food. Routledge, London.

Gelb, S. (Ed.) (1991) South Africa’s Economic Crisis. Zed Press, London.

Gertler, M. (1988) The limits to flexibility-comments on the post-Fordist vision of production and its geography, Transactions of the Institute of British Geographers, 13, 419-432.

Gertler, M. (1989) Resurrecting flexibility-a reply, Trans- actions of the Institute of British Geographers, 14, 109- 112.

Gertler, M. (1992) Flexibility revisited-districts, nation- states, and the forces of production, Transactions of the Institute of British Geographers, 17,259-278.

Goodman, D. and Watts, M. (1994) Reconfiguring the rural or Fording the divide? Capitalist restructuring and the global agro-food system, Journal of Peasant Studies, 22,149.

Humphrey, J. (1995) Industrial organization and manufacturing competitiveness in developing countries: introduction, World Development, 23, l-7.

IDC (1992) Sectoral Data Series-Manufacturing, Depart- ment of Economic Research and Development, IDC, Pretoria.


Nolan, P. and O’Donnell, K. (1987) Taming the market economy? an Assessment of the GLC’s experiment in restructuring for labour, Cambridge Journal of Econ- omics, 11, 2.51-263.

Piore, M. and Sabel, C. (1984) The Second Industrial Divide: Possibilities for Prosperity, Basic Books, New York.

RDP (1994) The Reconstruction and Development Pro- gramme: a Policy Framework. Umanyano Publications, Johannesburg.

Rogerson, C. (1994) Flexible production in the developing world: the case of South Africa, Geoforum, 25, 1-17.

Rustomjee, Z. (1992) The Boundaries of the Mineral- Energy-Complex: The Implications for Manufacturing Led Growth Strategies, In: Papers submitted for the Competition to Promote Academic Research into Indus-

119

trial Policy and Development in South Africa, L. van Pletsen and J. Hurter (Eds). IDC, Sandton.

Sabel, C. and Zeitlin, J. (1985) Historical alternatives to mass production: politics, markets and technology in nineteenth-century industrialisation, Past and Present, 108,133-176.

Sabel, C. (1982) Work and Politics. Cambridge University Press, Cambridge.

Schoenberger, E. (1989) Thinking about flexibility-a response to Gertler, Transactions of the Institute of British Geographers, 14, 98-108.

South African Statistics (1990) Central Statistical Services, Pretoria.

Union Statistics for Fifty Years (1960) Union Statistics for Fifty Years: Jubilee Issue, 1910-1960. Bureau of Census and Statistics, Pretoria.

Flexible production and flexible theory: the case of South Africa

Documents

Transcript of Flexible production and flexible theory: the case of South Africa